For the first time in the history of devolution, nationalist or independence-oriented parties lead all three devolved nations simultaneously: Sinn Féin in Northern Ireland, Plaid Cymru in Wales, and the SNP in Scotland. Meanwhile, Reform UK topped the national equivalent vote in England. The United Kingdom's political geography has been fundamentally redrawn. What has not been redrawn — and what makes the May 2026 results constitutionally as well as politically significant — is the legal framework within which each of these parties must now operate. Northern Ireland has a treaty-backed statutory mechanism for a referendum on its constitutional future. Scotland has a route that is blocked. Wales has a mechanism that has never been activated. These are not minor technical differences; they are the fault lines along which the UK's territorial constitution may be tested in the years to come.
Northern Ireland: A Statutory Pathway
Northern Ireland has a legally defined route to secession. The Belfast/Good Friday Agreement (GFA), recognised that it is for the people of the island of Ireland alone to exercise their right of self-determination to bring about a united Ireland, if that is their wish. Critically, the GFA distinguishes between the border poll mechanism and the full trigger for unification. A border poll in Northern Ireland returning a majority for unity does not on its own discharge the process: concurrent consent from the Republic of Ireland is also required. Only when that concurrent exercise of self-determination has occurred does the Agreement's binding obligation apply: paragraph 1(iv) provides that it will then be ‘a binding obligation on both Governments to introduce and support in their respective Parliaments legislation to give effect to that wish’.
Those commitments are reflected in UK domestic legislation. Section 1 of the Northern Ireland Act 1998 and Schedule 1 give the Secretary of State both a discretionary power to order a border poll and a conditional duty to do so if it appears likely that a majority would vote for a united Ireland — subject only to the constraint in Schedule 1, paragraph 3 that no poll may be ordered within seven years of a previous one. The High Court of Justice in Northern Ireland has clarified that this assessment must be based on an honest reflection of the available evidence — election results and opinion polls can form part of that evidential context.
Sinn Féin has been the largest party in Northern Ireland since 2022 and Michelle O'Neill has been First Minister since 2024. The Secretary of State has not yet determined that a border poll is required. But the legal machinery exists, is reasonably well-understood, and has been tested in litigation. Of the three devolved nations, Northern Ireland is in the constitutionally clearest position — at least as regards the formal mechanism for calling a poll.
That qualification matters. As Eustace has argued, the 2022 Independence Referendum Bill Reference from Scotland may have implications for Northern Ireland that are not yet fully worked through. The "relates to reserved matters" test applies equally to Acts of the Northern Ireland Assembly: if the Assembly were to legislate for measures whose purpose was to build political pressure toward a united Ireland — rather than simply exercising its ordinary devolved functions — it may face similar constraints to those now imposed on the Scottish Parliament. The Assembly has no formal role in triggering a border poll under Schedule 1 of the 1998 Act, and may also be limited in what it can do to promote one. The clarity of the Secretary of State's Schedule 1 powers is not matched by equivalent clarity about what a Sinn Féin-led Executive can lawfully do in the space below that threshold.
Scotland: A Blocked Route
Scotland's position is almost the inverse. The political case for a further independence referendum has been made repeatedly and successfully at the ballot box. The SNP has won consecutive elections on manifesto commitments to a second referendum, and the current Holyrood Parliament contains a pro-independence majority of 73 seats out of 129. And yet the legal route is, for now, closed.
The UK Supreme Court's 2022 ruling in the Independence Referendum Bill Reference held unanimously that any Bill providing for a referendum on Scottish independence would have ‘more than a loose or consequential connection’ with the reserved matters of the Union of the Kingdoms of Scotland and England and the Parliament of the United Kingdom, and was therefore ultra vires for the Scottish Parliament. The significance of that ruling is best understood alongside the two decisions that preceded it. In the Brexit Continuity Bill Reference [2018], the Court held that Scottish legislation making the effect of Westminster subordinate legislation conditional on Scottish Ministers' consent was outwith Holyrood’s competence, since it ‘would render the effect of laws made by the UK Parliament conditional on the consent of the Scottish Ministers' — a modification of section 28(7) of the Scotland Act, which preserves Parliament's unqualified legislative power. In the Incorporation Bills Reference [2021], the Court extended that reasoning to measures which, while not formally conditioning Westminster legislation, would nonetheless 'impose pressure on Parliament to avoid the opprobrium' of incompatibility findings. Together, these decisions under s.29(2)(c) establish that the Scottish Parliament cannot use its legislative competence to constrain Westminster's sovereign choices, directly or indirectly. The 2022 ruling, operating under the distinct but related ground of s.29(2)(b), applied the same logic to referendums: even non-self-executing legislation whose purpose is to generate democratic pressure on a reserved constitutional question has ‘more than a loose or consequential connection’ with that question. As Eustace has argued, the cumulative effect of all three judgments is that the devolved institutions are precluded not only from legislating for independence but from taking measures whose purpose is to pressure the UK Parliament into agreeing to the dissolution of the Union — a principle whose outer boundaries, as McHarg observed in her evidence to the Constitution, Europe, External Affairs and Culture Committee of the Scottish Parliament, remain unclear.
The only lawful route (other than primary legislation) established by precedent is the section 30 procedure: the UK Government may use the Crown's Order in Council power to modify Schedule 5, as it did under the Edinburgh Agreement of October 2012, with the Order itself following in 2013. This requires the consent of the UK Government, which successive administrations since 2016 have declined to give. That said, in October 2025, the current Labour Government signalled it ‘would take an SNP majority seriously’. The May 2026 result, returning the SNP as the largest party in a minority position, puts the precise application of that commitment in question.
I recently submitted evidence to the Scottish Parliament on whether any alternative trigger mechanism might be lawful. The Scottish Government's own position, set out in its September 2025 paper Your Right to Decide, is that an electoral mandate is itself sufficient: when the people of Scotland elect a Parliament with a clear pro-referendum majority following an explicit manifesto commitment, the UK Government is under a political obligation to negotiate a section 30 transfer. That paper accepts the UKSC judgment entirely but argues it does not extinguish the democratic right — it merely confirms that legal competence to act on it remains with Westminster. The difficulty with this position, as the paper itself implicitly acknowledges, is that it is vulnerable to UK Government refusal.
This is why a legal mechanism to reinforce the democratic mandate may be needed. The Committee itself, reporting in February 2026, concluded that steps should be taken to address this democratic anomaly as a matter of urgency. The proposal I put forward in the evidence is the following: the Scottish Parliament could arguably legislate for a referendum asking whether the Scottish Government should be mandated to enter negotiations with the UK Government for the issuing of a section 30 Order. Crucially, this question would not address the constitutional future of the Union — a reserved matter. It would instead merely ascertain whether the Scottish electorate wishes to mandate the Scottish Government to enter negotiations with the UK Government regarding the exercise of a power that the latter already possesses (section 30). The purported pressure from such a referendum would fall on the two governments to negotiate in good faith, not on the UK Parliament to legislate on the Union. On that basis, it is at least arguable that it would not have ‘more than a loose or consequential connection’ with the sovereignty of the UK Parliament or the Union— the Supreme Court's own test.
The argument is not without its problems, however. The UKSC's approach to purpose under section 29(3) of the Scotland Act is broad: the court asks not what the Bill says on its face but what it is ‘really about,’ drawing on manifestos, ministerial statements, and the full political context. In the Independence Referendum Bill Reference itself, the court cited the SNP's 2021 election manifesto as evidence of purpose. A consultative mandate referendum might be surrounded by identical contextual material, and a court could readily find that its real purpose was to build democratic pressure toward independence — precisely the kind of connection the cumulative jurisprudence has placed outside Holyrood's competence. More problematically, the Lord Advocate advanced a structurally similar argument in the Bill Referenceitself: since the Scottish Government can negotiate with the UK Government on reserved matters (including seeking a section 30 Order), the power to hold a referendum should extend to reserved matters. Notably, the court did not engage directly with this argument—though its overall conclusion at paragraphs [82]-[83] may implicitly foreclose the position.
In addition, the procedural obstacles are significant independently of the substantive legal question. Before any such Bill could be introduced, the Lord Advocate would need to clear the ministerial statement under section 31(1) of the Scotland Act that the Bill is within legislative competence — a threshold that, given the current state of the jurisprudence, they might find difficult to meet. If they did sign off, UK Law Officers could in principle seek a pre-introduction reference under paragraph 34 of Schedule 6; but the Court retains a discretion to decline such references and might well prefer the matter to proceed through the section 33 route following the Bill's passage, unless the circumstances were considered sufficiently exceptional to justify departure from the ordinary parliamentary process. At the same time, there is nothing that prevents the UK Parliament from legislating to clarify that holding such a referendum is outwith Holyrood’s competence. The argument is best understood, therefore, as a constitutional claim with genuine legal ingenuity — one that deserves to be tested — rather than a route whose lawfulness is guaranteed. Its greater force may be political: by constructing the most defensible lawful mechanism available and placing the onus on Westminster either to engage or to foreclose it by legislation, the Scottish Parliament would expose the constitutional asymmetry more sharply than any electoral mandate alone.
Wales: A Mechanism That Exists But Has Never Been Used
Wales appears at first glance to be in the most legally undeveloped position of the three. Plaid Cymru's historic victory raises the question of self-determination more acutely than ever. And yet, on closer inspection, the legal framework is less silent than it seems.
The Welsh equivalent of a section 30 Order is section 109 of the Government of Wales Act 2006. It is headed — identically to section 30 of the Scotland Act 1998 — ‘Legislative competence: supplementary’. It provides that the Crown may by Order in Council amend Schedules 7A and 7B (which define the Senedd's legislative competence and the reserved matters respectively). A section 109 Order could, in principle, modify the specific reservation in Schedule 7A that protects ‘the union of the nations of Wales and England,’ in the same way that the Edinburgh Agreement of 2012 used a section 30 Order to lift the analogous Scottish reservation enabling the independence referendum. Both provisions impose the same procedural requirement: no recommendation may be made to the Privy Council unless a draft has been approved by each House of Westminster and the devolved legislature. For Scotland, this obligation flows from Schedule 7 (in which section 30 is stated to require the ‘Type A’ procedure) read with section 115 of the Scotland Act. For Wales, it is expressly stated in section 109(4) itself.
This is important in light of the final report of the Independent Commission on the Constitutional Future of Wales, which assessed three options — enhanced devolution, a federal UK, and an independent Wales — and concluded that each is viable. It affirmed that the UK is a voluntary union of nations and that the people of Wales have the right to determine their constitutional future. But it was candid about what it had not resolved: it acknowledged that it had not had time to consider fully what processes should be followed to enable a referendum on Wales's constitutional future, and called for urgent consideration by both governments of the conditions under which such a referendum could and should be held. That call has gone unanswered. The May 2026 results make that call harder to defer.
The First Minister of Wales Rhun ap Iorwerth has already ruled out holding an independence referendum in a first term of government. But the legal question will not remain dormant indefinitely. If and when it is pressed, the approach I have proposed for Scotland is equally available in Wales. The same legal difficulties that attend the Scottish version apply here: a court applying the Government of Wales Act's equivalent legislative competence test would ask what such a referendum was ‘really about,’ and the answer in context might be difficult to separate from the question of Welsh independence itself.
The Case for Proceduralisation
The differences between the three frameworks reflect the different historical and political trajectories of the devolved nations. In that sense, the asymmetry between three constitutionally distinct situations is to a certain extent defensible. What is harder to defend is something more specific: that two of the three nations operate in a constitutional void in which consistent democratic mandates can be indefinitely blocked without any principled constitutional response. This is (as I have argued here and here), constitutionally counterproductive. The perceived advantage of managing secessionist claims within the rule of law — rather than leaving them as political facts that lie outside it — is that it transforms what might otherwise become a paralysing constitutional tug of war into a managed democratic process. A clear procedural channel for self-determination claims creates incentives for cooperation and compromise that outright refusal cannot provide.
This is the insight that the Quebec Reference contributes, though Canada's framework is itself contested. Quebec's National Assembly repudiated the Clarity Act's principles by passing Bill 99, asserting that the province alone could set the conditions for a referendum; secession referendums are not required under the amending formula; and the conditions and modalities under which the duty-to-negotiate would operate remain deeply contested between the federal and provincial levels. The value of the Quebec Reference lies not in providing a mechanism to replicate but in demonstrating that a constitutional order can articulate what Tancredi has called a "normative due process" for secession claims: a framework of constitutional principles — democracy, constitutionalism, the rule of law — that governs when and how competing parties are obliged to engage. The UK Supreme Court expressly endorsed the Canadian court's limits on external self-determination but did not engage with that dimension of the framework, leaving open whether the duty-to-negotiate logic has any resonance in the UK context.
Overall, what the UK constitutional order lacks is a set of constitutional principles that set out the democratic conditions that give rise to obligations on the UK Government to respond. The result is a constitutional order in which the question of whether the union remains voluntary can be raised repeatedly at the ballot box and blocked indefinitely by executive decision. Whether the UK Government chooses to develop a principled framework or whether Holyrood and the Senedd test the outer limits of their existing competences remains to be seen. The fault lines are drawn. Whether they are managed within the rule of law is one of the main questions for the years to come.
This article was first published on UKCLA Blog on 21 May 2026.